—  SPECIALTY CONFERENCE  —

Bone & Soft Tissue Pathology

Case 2 - Osteoblastoma with Secondary ABC

Michael J. Klein
University of Alabama Medical Center
Birmingham, AL


Click on each slide thumbnail image for an enlarged view
Clinical History
This 17-year-old male had begun to complain of pain just above the medial left knee about three months prior to hospital admission. The pain was first noticed at rest but became so severe that it limited his daily activities. It interfered with sleep, and it was not relieved by over the counter non-steroidal anti-inflammatory agents. On physical examination he was a normally developed teenager in obvious discomfort but not in distress. There was moderate to severe tenderness of the medial knee accompanied by a sensation of deep soft tissue fullness. Laboratory findings were not remarkable.


Case 2 - Figure 1 - Lateral and AP conventional radiographs demonstrate an ovoid, fairly defined lesion just cephalad to the medial femoral condyle. The lesion is more radiolucent in the middle than at the extremes imparting a somewhat beveled appearance. The AP view suggests that the cortex is focally interrupted medially and that there is an ill-defined soft tissue density outside of the cortex. There is an ovoid, smaller lesion medially with scalloped, sclerotic borders abutting the lateral femoral cortex in the metadiaphyseal region of the femur. In the lateral view, the lesions are superimposed and the scalloped, sclerotic edge of the lateral lesion is more easily seen. In addition, there is slight bulging of the cortex posteriorly
Case 2 - Figure 2 - A fat-suppressed coronal MRI demonstrates that the larger, medial lesion demostrates a mixed signal intensity. Although there is a relatively circumscribed outer edge with low signal, there is a radial high signal surrounding this edge, which may be lesional or edema. The medial edge of the lesion extends through the cortex into a mixed signal soft tissue mass which elevates the periosteum. The second lesion on the lateral side is multilocular and has a circumscribed low signal intensity.. The second coronal view on the left demonstrates the soft tissue mass and emphasizes the high signal about the lesion which extends into the end of the bone
Case 2 - Figure 3 - Two axial fat suppressed T2 weighted views of the distal femur reveal fluid levels in the soft tissue and intraosseous components of the medial lesion and a hypointense signal in the lateral lesion



Case 2 - Figure 4 - Four photomicrographs taken from the curettage at first operation demonstrate a lesion characterized by a very vascular spindle cell and giant cell containing solid tissue (lower left) in which there is a progression from very small complete vessels to partially complete large sinusoids containing blood in their lumens and in their walls (upper left and upper right). In addition, there is bone formation of the periosteal type appreciated at low power outside of the vascular tissue (upper right) as well as bone formation in the walls of the sinusoids (lower right)
Case 2 - Figure 5 - Four additional photomicrographs demonstrate solid areas in which giant cells and vascular tissue can just be appreciated (upper left) and, at higher magnification, the relatively even distribution of the giant cells in the solid areas (lower left) and the hypervascularity of the solid tissue bordering a sinusoid (lower right). The photograph at the upper right, though taken at low power, demonstrates bone formation of reactive type which is partly decalcified (center left) and on the right, more microtrabecular with hemorrhage and background hypercellularity. While the previous seven photographs were typical findings of aneurismal bone cyst, this area looks different at low magnification
Case 2 - Figure 6 - Higher magnification compares solid area of aneurismal bone cyst (lower half) with an area showing more microtrabecular bone with hypercelluarity seen in the last slide (top). Note that this bone is surrounded by appositional nuclei, which represent osteoblasts. In the right photograph, the field is entirely represented by osteoblastic, hypervascular tissue



Case 2 - Figure 7 - On the left is a wedge-shaped curetted fragment of purely osteoblastic tissue with small fragments of ABC. On the right, at high magnification is typical osteoblastoma composed of microtrabeculae surrounded by cells having the characteristics of osteoblasts. A few osteoclast like giant cells are present
Case 2 - Figure 8 - Last photograph at high magnification compares the features of the aneurismal bone cyst component, with an admixture of vessels, spindle cells and multinucleated giant cells (left) with that of osteoblastoma, showing trabecular bone/osteoid surrounded by pear-shaped osteoblasts and multinucleated giant cells (left)

Diagnosis: Osteoblastoma with Secondary ABC

Radiographic studies revealed at least two abnormalities of the distal femur. There was a large radiolucency with a slightly indistinct upper and lower boundary in the metaphyseal area that did not seem to extend past the area of the growth plate scar into the epiphysis. The medial cortex was somewhat indistinct and there was a soft tissue mass associated with a periosteal reaction that was not obviously continuous. The lateral metaphysis contained an eccentric, radiolucent lesion with scalloped, sclerotic borders centered in the cortex but probably also involving the cancellous bone. This lesion was elliptical in shape with its long axis parallel to the long axis of the femur. It was seen more distinctly in the AP views than in the lateral views where it overlapped with the larger medial lesion. The second lesion was diagnosed radiographically as fibrous cortical defect/non-ossifying fibroma.

Further imaging studies performed on the patient included an MRI that demonstrated that the soft tissue extension of the medial lesion was more circumscribed than originally demonstrable by conventional x-rays. In addition, the expanded extraosseous portion of this mass demonstrated fluid levels best seen on the axial T2 weighted images, while the signal remained low within the bone.

A clinical diagnosis of telangiectatic osteosarcoma was suspected and an open biopsy was performed (the section submitted is derived from one of four sections comprising all biopsy materials). Most of the sections contain an admixture of periosteal new bone, cartilage, and fibrous tissue suggesting a repair reaction. This repair is superimposed upon a background of tissue composed of spindle cells and multinucleated giant cells in which a hypervascular, spongy background demonstrating progression from very small capillary-like structures to large, blood-filled sinusoids sometimes without distinct endothelium. There is occasionally brisk mitotic activity in the spindle cell component, and scattered reactive osseous trabeculae. There is no nuclear atypia, pleomorphism, nor atypical mitotic activity. These findings are characteristic of aneurysmal bone cyst.

In only one of the four sections prepared, (the one handed out) there are two small solid islands of a tumor composed primarily of differentiating osteoblasts engaged in the bone production. This bone is wispy and microtrabecular in its orientation. There is no lace-like bone production, no sheet-like bone production, and no cartilage or fibrous tissue produced by the tumor. The tumor cells resemble osteoblasts, with eccentric, usually vesicular nuclei, sometimes with a prominent nucleolus, and there is no increase in nuclear/cytoplasmic ratio and no hyperchromatism. The cytoplasm of the tumor cells is basophilic to amphophilic. There is scant to minimal mitotic activity in the tumor cells, and no atypical mitoses are seen. Histologically, this tumor could be either osteoid osteoma or osteoblastoma, but it was diagnosed as osteoblastoma primarily because of its size on imaging, even though most of the lesion in these curettings was consistent with aneurysmal bone cyst.

The patient returned to the operating room the following week; a completion curettage revealed residual aneurysmal bone cyst. There was no evidence of residual osteoblastoma in six sections, which comprised the entire curetted material.

This case is interesting for several reasons. It is interesting radiographically, because while the clinical picture was thought to suggest a malignant primary bone tumor such as telangiectatic osteosarcoma, the process was in fact self-limited and benign. It is also interesting in that this patient had a distinct and separate benign lesion in the same bone that was not sampled in either of the curettings. Finally, it is interesting because the secondary aneurysmal bone cyst is so prominent a feature associated with a different primary benign bone tumor that it completely dominates that lesion radiographically and histologically.

Jaffe and Lichtenstein originally used the term osteoblastoma independently but Dahlin and Johnson had previously described the same lesion under the term "giant osteoid osteoma." Although it is identical histologically to osteoid osteoma, osteoblastoma has a different clinical history, skeletal location, and growth potential than osteoid osteoma. Osteoblastoma is usually painful, but the pain is not as disproportionately severe for its size as in osteoid osteoma. In addition, the pain of osteoblastoma is less often relieved by salicylates and other non-steroidal anti-inflammatory medications as it is with osteoid osteoma.

Osteoblastomas often arise in the axial skeleton. More than one-third of osteoblastomas are found in the neural arches of the vertebrae. While osteoid osteomas of the spine are associated with painful scoliosis, osteoblastomas often present with radiculopathies. This may be because osteoid osteomas are limited to 2 cm or less in diameter, whereas osteoblastomas are usuall 2-4 cm or more in diameter.

Radiographically, there is no absolutely typical pattern for osteoblastoma. It may appear as a circumscribed radiolucency like osteoid osteoma albeit larger. It may present as a "blown-out" lesion resembling aneurysmal bone cyst (even if no histological component of aneurysmal bone is identified). It may even (as in this case), be suggestive of a malignant neoplasm.

Histologically, Osteoblastoma consists of a well-vascularized fibrous stroma showing prominent osteoblastic differentiation. There are usually wispy immature bone trabeculae with a woven collagen fiber pattern, and these are often lined by osteoblasts. The cellular stroma even in areas where there is little bone formation demonstrates osteoblastic differentiation although the osteoblasts are usually not aggregated into solid sheets. Osteoblasts are pear-shaped with basophilic cytoplasm and often have enlarged, hyperchromatic nuclei. There are often osteoclasts remodeling trabeculae that are mineralized, but these are not necessary for the diagnosis. The diagnosis is more difficult if osteoblastoma reaches a size greater than 4 cm. and particularly, if it is complicated by one or more recurrences or has a locally more destructive pattern than typical osteoblastoma.

A subset of osteoblastomas has been termed aggressive osteoblastoma. Histologically, they contain fairly large numbers of so-called epithelioid osteoblasts, which are polyhedral cells that are at least twice the size of osteoblasts and have prominent nucleoli. In aggressive osteoblastoma, the osteoblasts not only line the osteoid trabeculae, but they may be disposed in sheets. It is probable that many lesions that have been reported under the name malignant osteoblastoma have been instances of particularly aggressive appearing osteoblastomas.

Occasionally, the osteoid formation in osteoblastoma is lace-like, and the mitotic activity is brisk. Since there are instances of osteoblastoma in which the radiographs are consistent with a malignant tumor and there are instances in which true osteosarcoma resembles osteoblastoma histologically, it is sometimes difficult to separate osteoblastoma from osteosarcoma. In general, the degree of nuclear atypia and mitotic rate is lower in osteoblastoma than in osteosarcoma and atypical mitoses are not found in osteoblastoma. The interface of lesional tissue and bone is very important diagnostically, because osteoblastomas are well circumscribed but osteosarcomas resembling osteoblastoma will often permeate the intertrabecular marrow spaces of normal adjacent bone. The presence of cartilage differentiation, although it has been rarely reported in osteoblastoma, should always raise the suspicion that the lesion is osteosarcoma until proven otherwise.

Jaffe and Lichtenstein first mentioned aneurysmal bone cyst (ABC) as a clinical entity in 1942 in their paper on solitary bone cysts. The term "aneurysmal" alluded to the fact that the lesion is expansile. Histologically, it is characterized by a spongiform, connective tissue stroma containing variably dispersed aggregates of osteoclast-like multinucleated giant cells. There are large numbers of vascular spaces arranged in a progression from very small capillaries to very large sinusoidal pools. Immunohistochemistry suggests that many of its vascular spaces are incompletely endothelialized, although its clinical behavior and MRI appearance do not agree with that assessment. Aneurysmal bone cyst tends to be rich in osteoblastic activity, and reactive bone having prominent appositional osteoblasts is sometimes so dominant near its sinusoidal walls that ABC can be misdiagnosed as osteoblastoma. This is an important point in the submitted case, because its histology contains diagnostic features of both lesions intimately admixed with one another.

The association of ABC with other bone lesions is a well-recognized phenomenon. It has been described most frequently in combination with the benign bone tumors chondroblastoma, giant cell tumor, osteoblastoma, and chondromyxoid fibroma. It also occasionally complicates non-ossifying fibroma, fibrous dysplasia, and other bone lesions, albeit less commonly. In some instances, the ABC reaction is discovered only on histological examination and not suspected clinically. In many instances, however, the radiographic appearance suggests ABC and the underlying lesion is discovered admixed with the ABC dominating the histologic sections. This type of ABC, sometimes referred to as secondary aneurysmal bone cyst, comprises about half the cases of ABC. Since de novo, or primary ABC tends to arise prior to the third decade, any lesion thought to be ABC in an individual over the age of twenty should be examined very carefully for the presence of any other underlying lesion.

While primary ABC may be mistaken for osteoblastoma because of reactive bone formation, ABC may also be mistaken for giant cell tumor because it can contain extensive fields of randomly dispersed giant cells in a fibrous stroma. If the patient is skeletally immature (if the growth plates are open), a diagnosis of giant cell tumor is for all practical purposes ruled out. If the patient is skeletally mature, a diagnosis of giant cell tumor complicated by ABC is more likely.

The precise assessment of ABC as a distinct disease type is still not settled. Dorfman and Czerniak's very useful definition is a modern modification of the original description by Jaffe and Lichtenstein; "A peculiar lesion of bone that is characterized by the presence of spongy or multilocular cystic tissue filled with blood. The process is benign in nature, but it is locally destructive and has a high propensity for recurrence. Microscopically, cystic spaces are bordered by septa composed of a well vascularized, loose, fibroconnective tissue with prominent giant-cell reaction and focal reactive bone formation…" Although this definition stops short of calling ABC a neoplasm, the clinical behavior it describes comes tantalizingly close.

Though we may not know for certain under what disease to classify ABC, we have learned much about its anatomy and physiology from our colleagues in Radiology. Perhaps most interesting about ABC is that despite the fact that it bleeds at the same rate as a transected vein if unroofed, MRI demonstrates that blood flow inside an intact ABC is incredibly sluggish. We know this because if a patient remains motionless during and MRI study, T2 weighted images in the axial and sagittal planes demonstrate fluid-fluid levels in the locules of the lesion. The fluid above the level is bright signal and that below the level is signal hypointense; what is happening is sedimentation of erythrocytes from blood plasma (the plasma is bright on T2 and the erythrocytes do not generate a signal). This sedimentation of erythrocytes would not take place if blood flow were brisk. It would also never happen if blood were clotted, or if previously clotted blood were defibrinated. The process requires whole blood without significant movement, like blood would behave in a glass tube with anticoagulant added. Since there is no anticoagulant in the locules of ABC, the best logical conclusion is that these spaces are lined by endothelium since blood not contained within intact endothelial spaces undergoes the clotting reaction. Although histological and immunochemical studies have not convincingly demonstrated that the vascular spaces of ABC are completely endothelialized, the lack of in vivo clotting and the development of fluid levels are the most convincing indirect evidence of this.

While the name aneurysmal bone cyst implies that it is both a cyst and a bone lesion, ABC may sometimes be neither. Although it was first described in a paper on simple cysts, ABC does not fulfill the pathological definition of a cyst (i.e., a fluid filled cavity lined by epithelium or a bladder or sac in the body). A paper by Sanerkin and associates in 1983 describing a series of bone lesions causing bone expansion that were histologically identical to the cellular solid areas of ABC without prominent sinusoids only made nomenclature matters worse. The term they used to describe their lesion, solid aneurysmal bone cyst, is an oxymoron if ABC is a true cyst, since the terms solid and cyst are mutually exclusive. To further complicate matters, lesions of soft tissues that are histologically identical to ABC but without having underlying osseous involvement have been more recently reported. If these lesions are etiologically identical to classical ABC, then even the term "bone" may also prove incorrect in nomenclature!

Sixty-three years after its first description, a unifying thread is beginning to emerge with the application of cytogenetic and molecular analyses to ABC. In 1999, Panoutsakapoulos et al. reported recurrent reciprocal translocations involving chromosomes 16 and 17 in ABC. Subsequent reports have found translocations between chromosome 17 and several other chromosomes. Most recently, Oliveira et.al. have linked gene rearrangements of the USP6 and CDH11 oncogenes in ABC using both molecular cytogenetic and molecular genetic confirmatory techniques. While these rearrangements were not universal (69% of 52 primary ABC's examined), there were two interesting and perhaps elucidating features. The first is that the gene rearrangements were present in six of the seven solid variants of ABC examined and in the single soft tissue example examined. Although the numbers may be small, this provides the first genetic link between classical, solid, and soft tissue ABC and other evidence besides hematoxylin and eosin sections, that they may be truly related. The second is that ofanother 17 so-called secondary ABC's tested, there were no CDH11or USP6 rearrangements detected. This makes it seem as though so-called secondary ABC, a well recognized occurrence that comprises 50% of ABC cases, may actually reresent a histologic mimicker of true ABC.

While the radiographs also demonstrate a non-ossifying fibroma in this case, the curetted tissue did not contain the lesion. Non-ossifying fibroma and fibrous cortical defects, though they are clinically distinct, are pathologically identical and some pathologists (I am one of these) consider them one and the same. Together, they constitute the most common space-occupying lesion of bone with an estimated incidence of 33-40% of growing individuals. For the most part, they are developmental and self-limited lesions that involute and heal, although the very rare large non-ossifying fibroma will occasionally undergo pathologic fracture. One in four individuals with fibrous cortical defect have them in more than a single bone. The most common site is in the most rapidly growing places in the fastest growing bones, namely, the distal femoral and proximal tibial metaphyses.

Certainly, one in three patients presenting with a primary bone tumor will harbor a non-ossifying fibroma/fibrous cortical defect in some other bone. Since the same sites that have the highest incidence of FCD/NOF are also the most common loci of malignant and benign primary bone tumors, it might be expected that non-ossifying fibroma should be found fairly often as a coincidental entity in the same bone containing another primary bone tumor. This is, in fact, not the case. There are only eight other instances of this association reported in the literature; the most common one thus far reported is non-ossifying fibroma and osteosarcoma.

Suggested readings

  1. Abdelwahab, IF, Klein,MJ, Kenan,S et al: Coexistence of Primary Bone Tumors: Report of 4 Cases of Collision Tumors. JACR 53(5) 296-302, 2002.
  2. Dahlin, D.C., Johnson,E.W., Jr.: Giant Osteoid Osteoma. JBJS ,36-A, 559, 1954)
  3. Dorfman,HD, Czerniak,B.: Bone Tumors (St.Louis:Mosby), 1998.
  4. Dorfman,HD, Weiss,SW: Borderline Osteoblastic Tumors: Problems in the Differential Diagnosis of Aggressive Osteoblastoma and Low-grade Osteosarcoma. Semin Diagn Pathol 1:215-234,1984.
  5. Jaffe, H.L., Lichtenstein, L.: Solitary Unicameral Bone Cyst. With Emphasis on the Roentgen Picture, the Pathologic Appearance, and the Pathogenesis. Arch surg 44:1004-1025, 1942.
  6. Klein, M.J.,Parisien,M.V.,Schneider-Stock,R.: Osteoid osteoma. In: Fletcher, C.D.M., Unni,K.K., Mertens, editors. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Soft Tissue and Bone. Lyon: IARC Press: 2002, 260-261.
  7. Malcolm,AJ,Schiller,A.L.,Schneider-Stock,R: Osteoblastoma. In: Fletcher, C.D.M., Unni,K.K., Mertens, editors. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Soft Tissue and Bone. Lyon: IARC Press: 2002, 262-263.
  8. Nielsen,G.P.,Fletcher, CDM, Smith,MA, Rybak,L, Rosenberg,AE: Soft Tissue Aneurysmal Bone Cyst: A Clinicopathologic Study of Five Cases. Am J Surg Pathol 26(1): 64-69, 2002.
  9. Oliveira, A.M., Perez-Atayde, A.R., Inwards,C.Y. et.al.: USP6 and CDH11 Oncogenes Identify the Neoplastic Cell in Primary Aneurysmal Bone Cyst and Are Absent in So-Called Secondary Aneurysmal Bone Cysts. Am Jour Pathol 165:1773-1780, 2004.
  10. Oliveira, A.M., Bae-Li, H., Weremowicz,S., et.al: USP6 (Tre2) Fusion Oncogenes in Aneurysmal Bone Cyst. Cancer Research 64:1920-1923, 2004.
  11. Panoutsakapoulos, G., Pandis, N, Kyriazoglou, I. et.al.: Recurrent t(16:17)(q22:p13) in aneurysmal bone cysts. Genes, Chromosomes, Cancer 26:265-266, 1999.
  12. Rodriguez-Peralto,JL, Lopez-Barea,F, Sanchez-Herrera, S., Atienza, M: Primary Aneurysmal Cyst of Soft Tissues (Extraosseous Aneurysmal Cyst). Am J Surg Pathol 18(6): 632-636, 1995.
  13. Rosenberg, A.,Nielsen,G.P., Fletcher, J.A.: Aneurysmal Bone Cyst. In: Fletcher, C.D.M., Unni,K.K., Mertens, editors. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of Soft Tissue and Bone. Lyon: IARC Press: 2002, 338-339.
  14. Sanerkin, NG, Mott,MG, Roylance,J: An Unusual Intraosseous Lesion with Fibroblastic, Osteoclastic, Osteoblastic, Aneurysmal and Myxoid Elements: "Solid" Aneurysmal Bone Cyst. Cancer51:2278-222286, 1983.